Full-Stack Engineer • NLP / ML Specialist • Open-Source Builder
I'm a full-stack engineer and NLP specialist focused on building production AI/ML systems. My core work is in natural language processing — RAG pipelines, LLM fine-tuning, evaluation harnesses, and conversational interfaces — alongside deep learning, computer vision, and embedded edge inference. I build low-latency voice interfaces with streaming STT/TTS, and I've shipped real-time edge AI on embedded hardware with quantized models running on NPU accelerators.
I've built enterprise AI assistants at Amelia.ai, ML delivery infrastructure at eBay, blockchain systems at Tilting Point (including rescuing $1M+ in MATIC), and edge AI surveillance systems on embedded SBCs shipping 5,000+ produced units. Before that, I worked at digital ad agencies building AR/VR and mixed reality experiences.
I created and open-sourced AgentOS, a modular orchestration runtime for autonomous AI agents with cognitive memory, guardrails, and a plugin system, and Wunderland, a CLI framework built on AgentOS for deploying agents across 28 channels with personality modeling and on-chain identity via Solana.
With a BFA in Game Design from Parsons (2017), I bring a background in game design, AR/VR, and interaction design that shapes how I approach UX, narrative, and conversational interfaces. That perspective drives my work at Manic.agency, where I publish technical writing at The Looking Glass.